#!/usr/bin/env python3
"""
Simple Hugging Face upload script for Bengali AI model
"""

import os
from huggingface_hub import HfApi, create_repo, upload_folder

def upload_model_simple():
    """Upload model with environment-based authentication"""
    
    print("🚀 Uploading Bengali AI to Hugging Face Hub...")
    print("=" * 50)
    
    # Initialize API
    api = HfApi()
    
    # Repository details
    repo_id = "megharudushi/Sheikh"
    local_dir = "./ready_bengali_ai"
    
    if not os.path.exists(local_dir):
        print(f"❌ Directory {local_dir} not found!")
        return False
    
    try:
        # Check authentication
        print("🔑 Checking Hugging Face authentication...")
        user_info = api.whoami()
        print(f"✅ Authenticated as: {user_info['name']}")
        
        # Create repository
        print(f"🔗 Creating repository: {repo_id}")
        repo_url = create_repo(
            repo_id=repo_id,
            exist_ok=True,
            repo_type="model"
        )
        print(f"✅ Repository ready: {repo_url}")
        
        # Upload files
        print(f"📤 Uploading {len(os.listdir(local_dir))} files...")
        upload_folder(
            folder_path=local_dir,
            repo_id=repo_id,
            commit_message="Complete Bengali AI model with tokenizer and configs"
        )
        
        print("🎉 Upload successful!")
        print(f"🌐 Your model: https://huggingface.co/{repo_id}")
        
        return True
        
    except Exception as e:
        print(f"❌ Upload failed: {e}")
        
        if "authentication" in str(e).lower():
            print("\n🔧 Authentication Issue:")
            print("1. Set your HF token: export HF_TOKEN=your_token_here")
            print("2. Or use: huggingface-cli login")
            print("3. Or set token in code: api.login()")
        
        return False

if __name__ == "__main__":
    # Create model card
    readme_content = """# Bengali AI Model - Sheikh

A Bengali (Bangla) language AI model for instruction following and text generation.

## Model Details
- **Base**: microsoft/DialoGPT-medium  
- **Language**: Bengali (Bangla)
- **Parameters**: 355M
- **Size**: 1.4GB
- **Data**: Alpaca Bangla dataset

## Usage
```python
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("megharudushi/Sheikh")
model = AutoModelForCausalLM.from_pretrained("megharudushi/Sheikh")

# Bengali query
input_text = "বাংলাদেশের রাজধানী কী?"
inputs = tokenizer.encode(input_text, return_tensors="pt")
outputs = model.generate(inputs, max_length=150)
response = tokenizer.decode(outputs[0], skip_special_tokens=True)
print(response)
```

## Features
- Bengali language understanding
- Educational content generation  
- Cultural knowledge responses
- General conversation support
"""
    
    with open("./ready_bengali_ai/README.md", "w", encoding="utf-8") as f:
        f.write(readme_content)
    
    print("📄 Model card created")
    
    # Attempt upload
    success = upload_model_simple()
    
    if success:
        print("\n🎉 SUCCESS! Your Bengali AI is now on Hugging Face!")
        print("Repository: https://huggingface.co/megharudushi/Sheikh")
    else:
        print("\n⚠️ Upload failed. Check authentication and try again.")